Device for phase-based shot analysis in sport shooting using dual inertial measurement units and correlation with electronic hit evaluation

Abstract

(Main claim): Device for analyzing shot quality in sport shooting, comprising: • at least two inertial measurement units (IMUs), each comprising at least one 3-axis accelerometer and one 3-axis gyroscope, wherein the at least two IMUs are arranged at two positions spaced apart from each other along the longitudinal axis of the firearm, wherein a first IMU is located near the grip and a second IMU is located near the barrel or muzzle, • a data interface to an electronic hit evaluation system (EST) that records the hit position as a coordinate on a target, • a processing unit configured to calculate a differential motion from the signals of at least two IMUs, where the differential motion is defined as the difference between the acceleration vectors and / or the rotation rate vectors of the at least two IMUs, and to divide the firing sequence into at least three temporal phases and evaluate each phase independently, • wherein the processing unit is further configured to derive a sighting error as a residual by correlating the phase rating with the hit position measured by the EST using an exclusion procedure, by calculating an expected scatter range from the phase ratings and diagnosing a hit position outside the expected scatter range with simultaneously high phase quality as a sighting error, • a feedback unit configured to display phase-specific feedback to the shooter on a display device after completion of the firing cycle.

Description

I. Technical Field

[0001] The present invention relates to a device for real-time analysis of shot quality in sport shooting, in particular in supported shooting with air rifles according to the rules of the International Shooting Sport Federation (ISSF) and the German Shooting Federation (DSB).

[0002] The invention combines two spatially separated inertial measurement units (IMUs) arranged at different positions along the barrel axis of a firearm with the hit data of an electronic target evaluation system to form an integrated analysis system that provides the shooter with real-time phase-specific feedback on the shooting sequence, including sight error diagnosis by means of a process of elimination. II. State of the art 11.1 Optical trainer systems (SCATT, Noptel, RIKA)

[0003] Optical training systems like the SCATT Expert, Noptel ST-2000, and RIKA Home Trainer use a weapon-mounted optical sensor that tracks the muzzle movement relative to a target. These systems capture the aiming action (trace) as an absolute position on the target. Their disadvantages are: they require a defined optical axis to the target, they do not function, or only function with limitations, with live ammunition on real targets, and they can only indirectly detect the trigger pull and follow-through via the resulting tracking deviation, not the actual force applied to the grip. Furthermore, these systems do not offer an open data interface and store data in proprietary, sometimes encrypted, formats. 11.2 IMU-based trainer systems (MantisX, HoldMaster)

[0004] The MantisX system (Mantis Tech, LLC, USA) uses a single inertial measurement unit (3-axis gyroscope and 3-axis accelerometer) that attaches to the weapon via a Picatinny rail or adapter. The system analyzes weapon movement before, during, and after trigger pull and provides feedback on trigger quality via a smartphone app (Bluetooth connection).

[0005] The disadvantages of the current state of the art in MantisX are: • Use of only a single IMU, which makes it impossible to differentiate between grip and muzzle position and to distinguish the cause of weapon movement (grip vs. external influence). • No connection to an electronic hit evaluation system. The system does not know the actual hit position and therefore cannot establish a correlation between sensor quality and hit result. • No possibility of diagnosing sighting errors, as no process of elimination is possible without hit data. • Insufficient sensor sensitivity for precision disciplines such as supported shooting, where weapon movements are extremely small. • Generic three-phase model (before / during / after the shot) that is not tailored to the specific requirements of individual disciplines. II.3 Electronic disc systems (DISAG, Meyton, SIUS, Intarso)

[0006] Electronic scoring systems (ESTs) such as DISAG OptiScore, Meyton, SIUS, and Intarso TrueScore record the hit position (X / Y coordinate) on a target using acoustic triangulation, optical light barriers, or camera-based methods with an accuracy of typically 0.1 mm or better. These systems provide the shooter with the ring value and the hit position, but offer no information about the shooting sequence, trigger quality, or possible causes of error. II.4 Known patents

[0007] US Patent 10,502,531 (Erange Corporation, 2019) describes a training system in which IMU data from sensors worn on the shooter's body are correlated with the shot result determined by image processing from a camera. This patent differs from the present invention in several essential aspects: (a) the IMU is placed on the body (wrist, compression sleeve), not on the firearm; (b) hit detection is performed by optical image analysis from a camera, not by a certified electronic target system; (c) it does not describe a dual IMU arrangement on the firearm with differential motion analysis; (d) it does not describe an elimination procedure for sight error diagnosis.

[0008] US patent US 2016 / 0033221 (Firearm Accessory) describes a weapon accessory with IMU, GPS and magnetometer for general motion detection, but without connection to a hit evaluation system, without dual sensor arrangement and without phase-based analysis. III. Object of the invention

[0009] The present invention is based on the objective of providing a device that enables the sport shooter to perform a comprehensive, causal analysis of the shooting sequence in real time, including the diagnosis of sighting errors that cannot be directly measured with conventional sensor systems.

[0010] In particular, the invention is intended to solve the following sub-problems: • Differentiation of the cause of a weapon movement (force on the grip by the shooter vs. movement of the entire weapon on the support) by spatially separate measurement at the grip and muzzle. • Real-time correlation of IMU sensor data with the hit data of an electronic target system to validate the sensor diagnostics. • Derivation of a parameter that cannot be measured directly (sighting error) by means of an exclusion procedure, in which the difference between the hit position expected from the sensor data and the actually measured position is evaluated. • Discipline-specific phase analysis of the shooting sequence that goes beyond the generic three-phase model of the state of the art. IV. Description of the invention IV.1 System overview

[0011] The system according to the invention consists of three main components (see below). Fig. 1): (A) Sensor unit with dual IMU arrangement: A sensor unit comprising at least two inertial measurement units (each comprising at least a 3-axis accelerometer and a 3-axis gyroscope) arranged at spatially separated positions along the barrel axis of the firearm. A first IMU (hereinafter referred to as the “grip IMU”) is located in the area of ​​the grip or trigger assembly. A second IMU (hereinafter referred to as the “muzzle IMU”) is located in the area of ​​the muzzle or barrel. Optionally, the sensor unit includes a MEMS microphone for acoustic shot detection. The sensor unit communicates wirelessly (WiFi and / or BLE) with a processing unit. (B) Electronic Hit Evaluation System (EST): An electronic target system that records the hit position as an X / Y coordinate with an accuracy of at most 0.5 mm. The EST can be an existing external system (e.g., DISAG OptiScore, Meyton, SIUS, Intarso TrueScore) whose hit data is transmitted to the processing unit via a data interface (network, serial, or proprietary). In a preferred embodiment, the EST is an integrated component of the overall system, with the processing unit performing both hit evaluation and sensor analysis on the same platform. Deep integration of the sensor unit and EST on a shared platform offers significant advantages for training applications compared to a simple data interface to an external EST: The sensor unit's analysis results—particularly the phase indicators, sight error diagnostics, and context-sensitive action instructions—are displayed directly on the same screen the shooter is already viewing the target image on. The shooter doesn't have to switch their gaze between two separate displays (target system and trainer app) but receives both the hit result and the causal analysis in a single, coherent view. This is crucial for training because, in prone shooting, the window for effective feedback is narrow, and switching between devices interrupts the immediate connection between body awareness and feedback.Furthermore, the deep integration enables bidirectional data flow control: Sensor analysis can enhance the representation of the hit image with sensor-based annotations (e.g., color-coding of shots with diagnosed sight errors in the hit image), and conversely, the EST can pass contextual information (e.g., series status, test shot marking) to sensor analysis to adapt the phase evaluation and action instructions depending on the situation. (C) Processing unit with analysis software: A processor (e.g., embedded computer, single-board computer, or server) that receives the raw IMU data from both sensors in a time-synchronized manner, calculates the differential movement, performs a phase analysis of the firing sequence, correlates the results with the EST hit data, and displays phase-specific feedback to the shooter on a display device after completion of the last analysis phase (follow-up) with a processing latency of no more than 200 ms. Complete feedback, including the follow-up assessment, is thus typically available within approximately 1200 ms after firing. Phases 1-3 (setup, aiming, trigger pull) can be preliminaryly assessed and displayed even during the ongoing firing cycle. IV.2 Dual IMU Arrangement and Differential Motion Analysis

[0012] The core of the invention is the arrangement of two IMUs at different positions along the barrel axis of the weapon (see...). Fig. 1, reference numbers 1 and 2). In an air pistol mounted on a support bar, the bar serves as a pivot point (lever). Forces acting on the grip (due to the shooter's hand, especially during the trigger pull) generate a movement that is first measurable at the grip IMU and, with a time delay and amplified, at the muzzle IMU. Conversely, forces at the muzzle (e.g., recoil) produce an inverse temporal pattern.

[0013] The difference motion Δ(t) between the two IMUs is defined as: Δa(t)=amundung(t)−agriff(t) Δω(t)=ωmouth(t)−ωhandle(t) where a denotes the acceleration vector and ω the rotation rate vector. For a rigid weapon at rest, Δ(t) = 0. Deviations from zero indicate relative motion, the time profile of which allows conclusions to be drawn about the cause: • Impulse first at the grip IMU: The malfunction originates with the shooter (trigger error, grip tension, premature release). The leverage of the support bar amplifies the effect towards the muzzle. • Simultaneous impulse at both IMUs: The entire weapon moves (e.g., impact against the rest, vibration of the shooting range). This disturbance is not caused by the shooter. • No significant difference, but hit offset: Both IMUs show a clean reading, but the hit is not in the expected area. This is an indicator of a sighting error (see IV.4). IV.3 Phase model of the shooting sequence

[0014] The invention defines a discipline-specific phase model that divides the shooting sequence into at least three, preferably four, successive phases (see Fig. 2, upper section). For supported shooting with an air pistol, the preferred model comprises four phases: Phase 1 - Setup (Stance Setup): Period from approximately 3000 ms to 1500 ms before firing. During this phase, the shooter picks up the weapon and places it on the support bar. The settling time (the decay of large movements) is measured. The quality is determined from the amplitude and decay rate of the acceleration signals from both IMUs. Phase 2 - Aiming Process: Period from approximately 1500 ms to 100 ms before firing. In this phase, the shooter aligns the weapon with the target and stabilizes their stance. Balance oscillations on the support bar (tilting, swaying) are measured. The differential movement Δ(t) indicates whether the instability originates from the hand (grip IMU) or from the weapon's balance (muzzle IMU). Phase 3 - Trigger Pull: Period from approximately 100 ms before to 50 ms after firing. During this phase, the trigger is pulled. The impulse signature of the trigger pull is measured. A clean trigger pull generates a symmetrical, short impulse signal. A jerked trigger pull generates an asymmetrical signal with pre-impulses (jerking the trigger). The dual IMU configuration allows for causal correlation: The trigger impulse is first and more strongly measurable at the grip IMU, as the force originates directly at the trigger finger. The amplified effect at the muzzle (lever effect) is captured by the muzzle IMU and shows the actual impact on the point of impact. Phase 4 - Follow-through: Period from approximately 50 ms to 1000 ms after firing. During this phase, the shooter is supposed to keep the weapon on target. The decay behavior after the shot is measured. Rapid decay indicates good follow-through. Slow decay with oscillations indicates premature release or tension release in the body. Here, too, the dual IMU allows differentiation between whether the movement originates from the shooter (grip first) or from the weapon mechanics (simultaneously).

[0015] Each phase is independently evaluated with a quality score (0-100%). The evaluation is based on the amplitude, frequency, and temporal structure of the acceleration and rotation rate signals from both IMUs, as well as their difference.

[0016] The aforementioned time specifications are exemplary guidelines for supported shooting with an air pistol. In a preferred embodiment, the phase boundaries are not defined by fixed time windows, but are derived adaptively from the signal characteristics of the IMU data. The transition from phase 1 to phase 2 is detected, for example, when the variance of the acceleration signal falls below a settling threshold. The beginning of phase 3 is detected by an increase in the jerk (time-dependent change in acceleration, da / dt) above a trigger threshold. The end of phase 4 is determined by falling below a variance threshold in the acceleration signal or by a maximum timeout. This signal-based phase detection allows adaptation to different shooter populations and disciplines without manual configuration of the time windows. IV.4 Sight error diagnosis by process of elimination

[0017] A key aspect of the invention is the derivation of a parameter that cannot be measured directly—the sighting error—by a process of elimination (residual analysis). The method is based on the following consideration: Hit position = f(Aiming process, trigger, follow-through, sight picture)

[0018] The first three variables (aiming, trigger pull, follow-through) are measured by the dual IMU array. The point of impact is measured by the electronic target system. The sight error is the only variable not directly measured and can therefore be determined as a residual (see below). Fig. 2, lower section): Sight error ≈ Actual hit position − Expected hit position (based on IMU data)

[0019] The process includes the following steps: • From the quality values ​​of phases 2-4, an expected maximum dispersion radius (rerwartungt) is calculated. The mapping of the quality values ​​to the expected dispersion radius is achieved, for example, using a weighted combination: rerwartungt = rmax × (1 - Qweighted), where Qweighted is the weighted average of the phase qualities (e.g., take-off 50%, targeting 30%, follow-up 20%) and rmax is the discipline-specific maximum dispersion radius (e.g., 8 mm at 10 m). For high quality values ​​(e.g., Qweighted > 0.8), rerwartungt is small (e.g., < 2 mm at 10 m). • The actual hit position (X / Y coordinate) is received by the EST and the distance is actually = √(x 2 + y 2 ) calculated to the center of the disc. • The residual R = ractual - rexpected is calculated. If R is significantly positive and the IMU phase quality was high (e.g., Q-weighted > 0.6), a sighting error is diagnosed. The significance threshold can be defined statically (e.g., R > 1.5 mm) or derived dynamically from the dispersion of the last N shots. • The direction of the residue is calculated as an angle α = atan2(y, x) from the hit coordinates and displayed to the shooter as a clock position (e.g. α = 60° ≙ 2 o'clock). • A pattern is analyzed over several shots in a series: Systematic offset in one direction indicates a consistent sighting error (e.g., incorrect head position), while random offset indicates an inconsistent sighting pattern. • Increasing sighting errors over the course of a series are interpreted as an indicator of fatigue of the head / neck muscles. IV.5 Real-time feedback system

[0020] After the follow-through phase is complete (approximately 1000 ms after firing), the processing unit generates the complete feedback with a processing latency of no more than 200 ms. This feedback is displayed on a screen (tablet, monitor) at the shooting range. Partial evaluations of phases 1-3 can be displayed progressively even during the ongoing firing cycle. The feedback comprises three levels of information: Level 1 - Result: Ring value and series status, taken from the EST. Level 2 - Sensor profile: Acceleration trace of both IMUs over the entire firing cycle, color-coded by phase, as well as four independent phase indicators (setup, aiming, trigger pull, follow-through) and a fifth indicator for sight error diagnosis. Level 3 - Action instruction: A context-sensitive, discipline-specific text instruction that, based on the phase analysis and sight error diagnosis, identifies the most likely cause of the error and suggests a specific correction.

[0021] The instructions follow a priority hierarchy: (1) sight error (diagnosis of exclusion), (2) fatigue trends over the series, (3) weakest IMU phase, (4) positive reinforcement if the sequence is clean. IV.6 Time synchronization and data flow

[0022] The temporal correlation of the IMU data to the EST hit data requires synchronization with an accuracy better than 1 ms. The sensor unit has a local clock (e.g., the microcontroller's internal oscillator) that provides monotonically increasing timestamps to the IMU samples. The processing unit synchronizes the sensor unit's local time base with the system time using one or more of the following methods: • A common time base is established via NTP (Network Time Protocol) or PTP (Precision Time Protocol) if both systems are connected over the same network. With a typical NTP accuracy of 1-10 ms, fine-tuning is achieved using the following methods. • Acoustic shot detection via a MEMS microphone in the sensor unit, which serves as a trigger event for timing. The timestamp of the microphone event in the sensor unit's local time base is used as an anchor point for mapping to the EST hit input. • Alternatively or additionally: Purely IMU-based shot detection by detecting a characteristic jerk peak (time derivative of acceleration, da / dt) and / or a rotation rate maximum (ω peak), which determine the shot time without additional sensors. • Alternatively or additionally: EST event-driven synchronization, in which the EST transmits a hit entry timestamp to the processing unit and the processing unit correlates this with the next shot detection event in the IMU data. • When combining multiple shot detection methods, a plausibility check is performed: A shot is only considered valid if at least two of the detectors (microphone, IMU jerk, EST input) trigger within a coincidence window (e.g. 50 ms).

[0023] To compensate for clock drift of the local oscillator of the sensor unit, a time synchronization message is periodically exchanged (e.g. every 60 s) between the sensor unit and the processing unit, from which a drift factor is calculated and applied to the IMU timestamps.

[0024] The IMU data is stored in a ring buffer with at least 4000 ms of history, so that when the shot detection signal arrives, the complete setup and aiming phase can be analyzed retrospectively. IV.7 Transferability to other disciplines and weapon types

[0025] The described principle of the invention is not limited to supported shooting with an air pistol. The basic principles – dual IMU difference measurement, phase-based analysis, and sight error diagnosis by process of elimination – can be transferred to other disciplines and weapon types, with the phase model being adapted to the discipline-specific movement sequences in each case. The essential adaptations for selected scenarios are outlined below.

[0026] Air rifle supported (LG supported, 10m): The phase model is almost identical to the described air pistol supported model. The rifle also rests on a support bar, the lever mechanism is comparable, and the four phases (setup, aiming, trigger pull, follow-through) are directly applicable. The key difference lies in the higher precision of the discipline: The inner 10-ring tolerance of the rifle is smaller (0.5 mm diameter), so the thresholds for phase evaluation and the elimination process must be calibrated accordingly more tightly. The muzzle IMU is positioned on the longer barrel of the rifle, which increases the leverage effect and enhances measurement sensitivity.

[0027] Air pistol and air rifle freehand (10 m): In freehand shooting, the support bar is eliminated as a pivot point. The weapon is held entirely by the shooter, resulting in a significantly greater range of motion than in supported shooting. The phase model is adapted as follows: Phase 1 ("Attack") encompasses raising the weapon and assuming the shooting position. Phase 2 ("Settling / Stabilization") encompasses the stabilization of the natural point of aim. Phases 3 and 4 (trigger release and follow-through) remain essentially unchanged. Dual IMU differential measurement is particularly valuable here, as it can distinguish whether a weapon movement originates from the shoulder-arm chain (uniform translation of both IMUs) or from a local force at the grip (differential movement). The process of elimination for sight error diagnosis is fully applicable, since the fundamental equation (point of impact as a function of aiming, trigger release, follow-through, and sight picture) is discipline-independent.

[0028] Small-bore firearms (25 m / 50 m): In small-bore disciplines (e.g., 25 m sport pistol, 50 m three-position rifle), recoil is an additional factor. This generates a characteristic impulse signal in both IMUs, which occurs after the trigger pull. The phase model is extended to include a "recoil management" phase, which analyzes the transition between firing and recoil absorption. Shot detection via the MEMS microphone is adjusted to the higher sound intensity. In rapid-fire disciplines (e.g., 25 m rapid-fire pistol, 5 shots in 4 seconds), the rhythm between successive shots is also analyzed, with the follow-through phase of one shot transitioning into the build-up phase of the next.

[0029] Large-caliber firearms (LCW): The method is generally applicable to large-caliber firearms (pistol, revolver, rifle). The stronger recoil simplifies shot detection and generates clearer signals in the IMU, thus increasing the measurement accuracy of the follow-through phase. The sensor unit is mounted via existing Picatinny rails or weapon-specific adapters. For large-caliber disciplines in hunting and law enforcement shooting, the phase models must be adapted to shorter cycle times and dynamic shooting positions (standing, kneeling, prone).

[0030] Dynamic disciplines (IPSC, PPC, multi-distance): In dynamic disciplines like IPSC (International Practical Shooting Confederation), multiple shots are fired at different targets at varying distances and positions within a short period. The phase model is fundamentally altered here: The classic setup and aiming phase is replaced by "target transition" (switching between targets), "target acquisition" (target acquisition), and "shot sequence" (firing sequence). Dual IMU measurement is particularly well-suited for analyzing the quality of target transitions (overshoot, stopping behavior). Integration with an electronic target system (EST) is feasible in this scenario via electronic steel targets (hit / no-hit and hit time) or camera-based evaluation systems.

[0031] All the aforementioned disciplines share the common feature that the three core elements of the invention – (i) dual IMU differential measurement for causally attributing the causes of malfunctions, (ii) correlation with an EST for validation, and (iii) the exclusion procedure for sight error diagnosis – are applicable regardless of the specific discipline. Adaptation is made only in the phase model (number, duration, and definition of the phases), in the threshold values ​​(calibration to the discipline-specific precision requirements), and in the mounting mechanism of the sensor unit on the respective weapon. V. Preferred Implementation Example V.1 Hardware

[0032] In a preferred embodiment, the sensor unit comprises: • An ESP32-S3 type microcontroller with integrated WiFi and BLE. • Two high-resolution IMUs of type ICM-42688-P (TDK InvenSense) or LSM6DSO (STMicroelectronics) with a sampling rate of at least 1 kHz, connected via SPI bus. • A MEMS microphone for acoustic gunshot detection. • A 3D-printed housing with a mounting mechanism for shooting air pistols (clamp mount or magnetic attachment). • The first IMU is positioned in the area of ​​the trigger guard / grip (distance to the support bar approx. 0-50 mm). The second IMU is positioned near the muzzle (distance to the support bar approx. 150-250 mm, depending on the weapon geometry).

[0033] The processing unit is a single-board computer (e.g. Raspberry Pi) on which server software (e.g. based on Rust / Axum) receives the IMU data via UDP / WebSocket, performs the phase analysis in real time and transmits the results to a tablet display at the shooting range via Server-Sent Events (SSE).

[0034] The electronic scoring system is an existing, certified EST (e.g. DISAG OptiScore, Meyton, SIUS, Intarso TrueScore or an open-source system) whose hit data is read by the processing unit via a network or serial interface. V.2 Signal Processing

[0035] The raw data from both IMUs are processed using a sensor fusion algorithm (e.g., Madgwick or Mahony filter) to determine orientation (roll, pitch, yaw) and linear acceleration. The differential motion is calculated as the vector difference between the linear acceleration and the angular velocity of both IMUs. Shot detection is achieved through threshold detection on the microphone signal, correlated with an acceleration maximum in the IMU data. V.3 Calibration

[0036] The system supports two calibration modes: • Fixed thresholds: Predefined quality thresholds based on the discipline (e.g., LP 10m prone shooting). Suitable for beginners. • Adaptive calibration: In an initial calibration phase (e.g., 20 shots), shooter-specific baseline values ​​are determined. The quality assessment is then performed relative to this individual baseline. This takes into account different performance levels (beginner vs. elite athlete). VI. Advantages over the state of the art • Causal analysis: The dual IMU arrangement allows the cause of a weapon movement (shooter vs. external influence) to be distinguished, which is not possible with a single IMU. • Sight fault diagnosis: By using the process of elimination in combination with the EST, a sight fault can be diagnosed that is not directly measurable with any known system. • Complete shot analysis: The combination of discipline-specific phase analysis (IMU) and hit result (EST) enables a comprehensive analysis of all relevant error factors in the shot. • Real-time capability: Complete feedback, including follow-through assessment, is displayed within approximately 1200 ms of firing. The progressive display of earlier phases during the firing cycle enables operant conditioning – the shooter can immediately correlate the feedback with their body awareness and train their muscle memory. • Discipline-specific: The phase model is tailored to the specific requirements of supported shooting and can be extended for other disciplines (freehand, three-position, rifle) with adapted phase models. • Longitudinal data acquisition: The persistent storage of all sensor and hit data enables the analysis of fatigue patterns, training progress and performance development over weeks and months. • Deep integration with EST: By integrating sensor analysis and hit evaluation on a single platform, the need to switch between separate devices is eliminated. The shooter receives the hit image and causal analysis in a cohesive view, ensuring a direct link between body awareness and feedback. Bidirectional data flow control enables sensor-based annotations in the hit image as well as context-sensitive adaptation of the sensor analysis to the competition or training status. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] US 10,502,531

[0007] US 2016 / 0033221

[0008] Cited non-patent literature

[0000] Optical trainer systems such as SCATT Expert, Noptel ST-2000

[0003]

Claims

[1] (Main claim): Device for analyzing shot quality in sport shooting, comprising: • at least two inertial measurement units (IMUs), each comprising at least one 3-axis accelerometer and one 3-axis gyroscope, wherein the at least two IMUs are arranged at two positions spaced apart from each other along the longitudinal axis of the firearm, wherein a first IMU is located near the grip and a second IMU is located near the barrel or muzzle, • a data interface to an electronic hit evaluation system (EST) that records the hit position as a coordinate on a target, • a processing unit configured to calculate a differential motion from the signals of at least two IMUs, where the differential motion is defined as the difference between the acceleration vectors and / or the rotation rate vectors of the at least two IMUs, and to divide the firing sequence into at least three temporal phases and evaluate each phase independently, • wherein the processing unit is further configured to derive a sighting error as a residual by correlating the phase rating with the hit position measured by the EST using an exclusion procedure, by calculating an expected scatter range from the phase ratings and diagnosing a hit position outside the expected scatter range with simultaneously high phase quality as a sighting error, • a feedback unit configured to display phase-specific feedback to the shooter on a display device after completion of the firing cycle. [2] Device according to claim 1, characterized by , that the processing unit is configured to infer the cause of the weapon movement from the temporal sequence in which a motion impulse is measured at the grip-adjacent IMU and at the barrel- or muzzle-adjacent IMU, whereby an impulse measured first at the grip-adjacent IMU indicates a disturbance caused by the shooter and a simultaneous impulse at both IMUs indicates an external influence. [3] Device according to claim 1 or 2, characterized bythat the processing unit is configured to adaptively derive the phase boundaries from signal characteristics of the IMU data, in particular by monitoring variance thresholds, jerk values ​​(time-dependent change in acceleration) and / or frequency band analysis, wherein the phases in an embodiment for supported shooting include: stance setup, aiming process, trigger pull and follow-through. [4] Device according to any one of the preceding claims, characterized by , that the processing unit is configured to analyze a pattern of diagnosed sight errors over several shots in a series, diagnosing a systematic offset in one direction as a consistent sight error, an increasing offset as a fatigue indicator, and a randomly distributed offset as an inconsistent sight image. [5] Device according to any one of the preceding claims, characterized by, that the feedback unit is configured to implement a priority hierarchy of instructions, with sight failure diagnoses taking precedence over phase-specific diagnoses, and with the feedback including a context-sensitive, discipline-specific instruction that identifies the most likely cause of the failure. [6] Device according to any of the preceding claims, characterized by that the electronic hit evaluation system and the sensor analysis are integrated on a common processing platform, wherein the analysis results of the sensor unit including phase indicators, sight error diagnosis and action instructions are displayed on the same display device on which the hit image is shown, and wherein the display of the hit image is enhanced with sensor-based annotations and contextual information from the EST is passed to the sensor analysis. [7] Device according to any of the preceding claims, characterized by , that shot detection is carried out by at least one of the following methods: (i) acoustic detection using a MEMS microphone, (ii) detection of a jerk peak and / or rotation rate maximum in the IMU data, (iii) reception of a hit input event from the EST, whereby, if multiple detectors are available, plausibility is verified by coincidence testing within a defined time window. [8] Device according to any one of the preceding claims, characterized by that the at least two IMUs each have a sampling rate of at least 200 Hz and are connected via a digital bus (SPI and / or I²C). 2 C) are connected to a microcontroller of the sensor unit, and that the sensor unit communicates with the processing unit via a wireless interface (WiFi and / or BLE) and / or a wired interface (USB). [9] Device according to any one of the preceding claims, characterized by , that the processing unit is configured to determine shooter-specific base values ​​for phase evaluation and expected dispersion radius in an initial calibration phase over a defined number of shots (e.g. 10-30), so that the quality evaluation is relative to the individual performance of the shooter. [10] Device according to any one of the preceding claims, characterized by , that the processing unit is configured to store the IMU raw data in a ring buffer with a history of at least the longest expected shot cycle, so that when the shot detection signal arrives, the complete setup and aiming phase can be retrospectively analyzed. [11] Device according to any of the preceding claims, characterized by, that the processing unit is configured to calculate the direction of the diagnosed sight error from the hit coordinates as angle α = atan2(y, x) and to display it to the shooter as a clock position on the display device. [12] Device according to any one of the preceding claims, characterized by that the device is configurable for different disciplines of sport shooting, wherein the phase model (number, duration and definition of the phases), the thresholds for the phase evaluation and the parameters of the exclusion procedure are adaptable to the respective discipline, in particular for supported shooting, freehand shooting, small-bore and large-bore disciplines as well as dynamic disciplines.

Images